"Traffic, Driving Style and Road Surface Condition" Dataset EDA

Content

Data Preparation

There are issues in the data types of all attributes but roadSurface, traffic and drivingStyle (Target)

Data Wrangling

Data Analysis

Here, we see that VehicleSpeedAverage along with EngineLoad may give a great information gain on the status of traffic. In addition, IntakeAirTemperature may be an indicator of traffic, however this needs to be studied whether it is affected by vehicle type or time of readings

Also FuelConsumptionAverage may need logarithmic transformation to show clearer data

Indicators for drivingStyle are harder to pinpoint by eye

Indicators for roadSurface are harder to pinpoint by eye

Summary of all correlations between attributes

Conclusion

This dataset explores multiple attributes of 2 different vehicle types and their relation to traffic (low, mid or high), drivingStyle (aggresive or normal) and roadSurface (even or uneven). It is an interesting dataset since it can be used to predict these mentioned 3 targets contributing to safer transportation since these metrics can be used to warn drivers in real time of surrounding dangers and alert the user of their own driving style

Pros:

Cons: